The HostAtlas Queue Monitor is a Laravel composer package that pushes queue metrics — pending jobs, failed jobs, processing jobs — to the HostAtlas platform. Queue health then shows up alongside the rest of your infrastructure in the same dashboard.

The package speaks to a generic /api/v1/queue/push endpoint that any language or framework can hit; the Laravel package just handles the auto-detection and scheduling for you.

Requirements

• PHP 8.2+
• Laravel 11 or 12
• A HostAtlas account with an API key

Install

composer require hostatlas/queue-monitor

The service provider is auto-discovered — no manual registration needed.

Configure

Publish the config file:

php artisan vendor:publish --tag=hostatlas-queue-config

Add to your .env:

HOSTATLAS_URL=https://my.hostatlas.app
HOSTATLAS_API_KEY=ha_your_api_key_here
HOSTATLAS_SERVER=web-prod-01

Variable	Description	Default
HOSTATLAS_URL	Your HostAtlas instance URL	https://my.hostatlas.app
HOSTATLAS_API_KEY	API key from Settings → API Keys	—
HOSTATLAS_SERVER	Server hostname — must match a server registered in HostAtlas	System hostname

Queue selection

By default, all queues are auto-detected. To monitor a specific subset, edit config/hostatlas-queue.php:

'queues' => ['default', 'emails', 'notifications'],

Use ['*'] to auto-detect every queue on the driver (default).

Usage

Scheduler (recommended)

Add to routes/console.php:

Schedule::command('hostatlas:queue-metrics')->everyMinute();

The scheduler-driven mode is the recommended production setup — HostAtlas expects a push per minute.

Manual

php artisan hostatlas:queue-metrics

Example output:

  default: 42 pending, 2 failed
  emails: 8 pending, 0 failed
  Queue metrics reported to HostAtlas.

Supported queue drivers

Driver	Auto-detect Queues	Metrics
Redis	Yes (scans Redis keys for queue names)	Size, failed count, system type
Database	Yes (GROUP BY queue on the jobs table)	Size, failed count, processing count
SQS	No (uses the configured queue name)	Size

What gets reported

Per queue:

• Queue name — auto-detected from the driver, or configured explicitly
• Pending jobs — current queue size
• Failed jobs — count from the failed_jobs table
• Processing jobs — jobs currently being worked (database driver only)
• System type — always laravel for this package

API push endpoint

The package POSTs to {HOSTATLAS_URL}/api/v1/queue/push with Authorization: Bearer <api-key>. You can push the same shape from any other stack:

curl -X POST https://my.hostatlas.app/api/v1/queue/push \
  -H "Authorization: Bearer ha_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "server": "web-prod-01",
    "queues": [
      {"name": "default", "size": 42, "failed": 2, "processing": 3, "system": "laravel"}
    ]
  }'

The system field is the free-form label HostAtlas uses to group metrics — pick one per stack:

• laravel (this package)
• bullmq (Node.js)
• sidekiq (Ruby)
• celery (Python)
• Anything else you’d like grouped together in the dashboard

Tip

The HostAtlas dashboard groups queue metrics by system, so you can see “Laravel queues” and “BullMQ queues” side-by-side across the fleet.

Multi-system support

The HostAtlas queue dashboard is stack-agnostic — Laravel is one supported system, but everything drives through the same push endpoint. Reference clients for the common ecosystems:

• Laravel — this package
• BullMQ (Node.js) — via API push (no first-party package yet)
• Sidekiq (Ruby) — via API push
• Celery (Python) — via API push

Security

• API key transmitted via Authorization: Bearer header over HTTPS only
• No sensitive data collected — only queue names and counts
• 10-second HTTP timeout per push request
• Errors are reported via Laravel’s report() helper — logged, never thrown, so a HostAtlas outage cannot break your app’s request path

Related

• CLI — hostatlas queues lists queue-monitor data from the terminal